markdownator 0.1.1 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 4799d3266ce18fa6adff8a3264b255a8bcaa2974888e260542adacd709566f2a
-  data.tar.gz: 50f573d19ff4b5407220fe4e8d2b50ebbc481c7d983183e24032ed6a8e4671f4
+  metadata.gz: ef75f03777f577049069b98fee3104ee31f7ad3d4cece96f3cc78052db648f6e
+  data.tar.gz: '047778687f87ca470f627d082e7653c17999c8981f7c579512eed0c8a31cb9fa'
 SHA512:
-  metadata.gz: 71873a123d242b1ff45147fc28eb50300c2dfb82d92dd5939be352dbeb05bbf9ae97c1c7003c5edf606e918176f5fa2dcb53b3f2ef2b0c3e9437a3fdb3faad81
-  data.tar.gz: 2c7aaf6871850fa39e323e32f7da9fe3870e06159fda18e4f04e268231214efd717c88f0cd41d291a07964b705e653cac677459a1921a1258545a68e72fd273e
+  metadata.gz: e3ce45e8c1f63c4f374aa475a27c3a567846423005caafa8999e136531139a36695fd180d806dcd5e49d1a5414b7a372d6a448f3c1bccb49d5868c13ae2e3870
+  data.tar.gz: ef73a6ccac7125b16fd8838a88f56be01c8fa555b39887963e8a79170c34428c21f20ac7fea867af06c6aa3410114ad01619f804ed3f1333669596bd656aed2c

data/CHANGELOG.md

@@ -7,6 +7,16 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.1.2] - 2026-06-13
+
+### Changed
+
+- XLSX conversion now reads the workbook directly with rubyzip and Nokogiri
+  instead of `roo`, so every Office format (DOCX, XLSX, PPTX, EPUB) shares one
+  approach and the `roo` dependency is dropped.
+- Moved the HTML renderer to `Markdownator::Renderers::HtmlRenderer` (it renders
+  Markdown, it does not convert a source file).
+
 ## [0.1.1] - 2026-06-13
 
 ### Changed
@@ -29,6 +39,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   dependencies.
 - Pluggable LLM image-captioner hook (off by default).
 
-[Unreleased]: https://github.com/alexrupom/markdownator/compare/v0.1.1...HEAD
+[Unreleased]: https://github.com/alexrupom/markdownator/compare/v0.1.2...HEAD
+[0.1.2]: https://github.com/alexrupom/markdownator/compare/v0.1.1...v0.1.2
 [0.1.1]: https://github.com/alexrupom/markdownator/compare/v0.1.0...v0.1.1
 [0.1.0]: https://github.com/alexrupom/markdownator/releases/tag/v0.1.0

data/Gemfile

@@ -17,5 +17,4 @@ gem "rubocop", "~> 1.21"
 gem "exifr", "~> 1.3"
 gem "nokogiri", "~> 1.15"
 gem "pdf-reader", "~> 2.12"
-gem "roo", "~> 2.10"
 gem "rubyzip", "~> 2.3"

data/README.md

@@ -16,7 +16,7 @@ libraries **lazily**, so you only install the gems for the formats you actually
 | HTML | `.html`, `.htm` | `nokogiri` |
 | XML | `.xml` | `nokogiri` |
 | Word | `.docx` | `rubyzip`, `nokogiri` |
-| Excel | `.xlsx` | `roo` |
+| Excel | `.xlsx` | `rubyzip`, `nokogiri` |
 | PowerPoint | `.pptx` | `rubyzip`, `nokogiri` |
 | PDF | `.pdf` | `pdf-reader` |
 | EPUB | `.epub` | `rubyzip`, `nokogiri` |
@@ -36,9 +36,8 @@ Then add the gems for the formats you need, e.g.:
 
 ```ruby
 gem "pdf-reader"      # PDF
-gem "roo"             # XLSX
-gem "rubyzip"         # DOCX, PPTX, EPUB, ZIP
-gem "nokogiri"        # HTML, XML, DOCX, PPTX, EPUB
+gem "rubyzip"         # DOCX, XLSX, PPTX, EPUB, ZIP
+gem "nokogiri"        # HTML, XML, DOCX, XLSX, PPTX, EPUB
 gem "exifr"           # image EXIF
 ```
 

data/lib/markdownator/converters/html.rb

@@ -1,6 +1,6 @@
 # frozen_string_literal: true
 
-require_relative "html_renderer"
+require_relative "../renderers/html_renderer"
 
 module Markdownator
   module Converters
@@ -20,7 +20,7 @@ module Markdownator
         Markdownator.require_optional("nokogiri", feature: "HTML conversion")
         doc = Nokogiri::HTML(html)
         root = doc.at_css("body") || doc.root || doc
-        HtmlRenderer.new.render(root)
+        Renderers::HtmlRenderer.new.render(root)
       end
 
       def self.extract_title(html)

data/lib/markdownator/converters/xlsx.rb

@@ -1,11 +1,12 @@
 # frozen_string_literal: true
 
-require "tempfile"
-
 module Markdownator
   module Converters
     # Converts an Excel .xlsx workbook into Markdown: one `## SheetName` heading
-    # and a Markdown table per sheet, using the `roo` gem.
+    # and a Markdown table per sheet.
+    #
+    # Reads the OOXML zip directly with rubyzip and Nokogiri, the same approach
+    # used by the DOCX, PPTX, and EPUB converters.
     class Xlsx < Base
       def accepts?(_io, stream_info)
         matches?(
@@ -16,12 +17,16 @@ module Markdownator
       end
 
       def convert(io, _stream_info, **_options)
-        Markdownator.require_optional("roo", feature: "XLSX conversion")
+        Markdownator.require_optional("zip", feature: "XLSX conversion")
+        Markdownator.require_optional("nokogiri", feature: "XLSX conversion")
 
-        with_tempfile(io) do |path|
-          workbook = Roo::Excelx.new(path)
-          sections = workbook.sheets.map { |name| render_sheet(workbook, name) }
-          Result.new(markdown: sections.compact.join("\n\n"))
+        ::Zip::File.open_buffer(io) do |zip|
+          shared = shared_strings(zip)
+          sections = sheets(zip).filter_map do |name, path|
+            table = markdown_table(parse_sheet(read(zip, path), shared))
+            "## #{name}\n\n#{table}" unless table.empty?
+          end
+          return Result.new(markdown: sections.join("\n\n"))
         end
       rescue StandardError => e
         raise FileConversionError, "Could not read XLSX: #{e.message}"
@@ -29,23 +34,85 @@ module Markdownator
 
       private
 
-      def render_sheet(workbook, name)
-        sheet = workbook.sheet(name)
-        rows = (1..sheet.last_row.to_i).map do |r|
-          (1..sheet.last_column.to_i).map { |c| sheet.cell(r, c) }
+      def read(zip, path)
+        zip.find_entry(path)&.get_input_stream&.read
+      end
+
+      # Ordered [[sheet_name, worksheet_path], ...] resolved via the workbook
+      # relationships.
+      def sheets(zip)
+        workbook = parse(read(zip, "xl/workbook.xml"))
+        return [] if workbook.nil?
+
+        rels = relationships(zip)
+        workbook.xpath("//sheets/sheet").filter_map do |sheet|
+          path = resolve_target(rels[sheet["id"]])
+          [sheet["name"].to_s, path] if path
         end
-        table = markdown_table(rows)
-        table.empty? ? nil : "## #{name}\n\n#{table}"
       end
 
-      def with_tempfile(io)
-        Tempfile.create(["markdownator", ".xlsx"]) do |file|
-          file.binmode
-          file.write(io.read)
-          file.flush
-          yield file.path
+      def relationships(zip)
+        doc = parse(read(zip, "xl/_rels/workbook.xml.rels"))
+        return {} if doc.nil?
+
+        doc.xpath("//Relationship").to_h { |rel| [rel["Id"], rel["Target"]] }
+      end
+
+      def resolve_target(target)
+        return nil if target.nil? || target.empty?
+
+        target = target.delete_prefix("/")
+        target.start_with?("xl/") ? target : "xl/#{target}"
+      end
+
+      # The shared string table: index -> text.
+      def shared_strings(zip)
+        doc = parse(read(zip, "xl/sharedStrings.xml"))
+        return [] if doc.nil?
+
+        doc.xpath("//si").map { |si| si.xpath(".//t").map(&:text).join }
+      end
+
+      def parse_sheet(xml, shared)
+        doc = parse(xml)
+        return [] if doc.nil?
+
+        doc.xpath("//sheetData/row").map do |row|
+          values = {}
+          width = 0
+          row.xpath("./c").each_with_index do |cell, position|
+            column = column_index(cell["r"]) || (position + 1)
+            width = column if column > width
+            values[column] = cell_value(cell, shared)
+          end
+          (1..width).map { |i| values[i] || "" }
+        end
+      end
+
+      def cell_value(cell, shared)
+        case cell["t"]
+        when "s" then shared[cell.at_xpath("./v")&.text.to_i].to_s
+        when "inlineStr" then cell.xpath("./is//t").map(&:text).join
+        when "b" then cell.at_xpath("./v")&.text == "1" ? "TRUE" : "FALSE"
+        else cell.at_xpath("./v")&.text.to_s
         end
       end
+
+      # Converts a cell reference like "B2" into a 1-based column index (2).
+      def column_index(ref)
+        letters = ref.to_s[/\A[A-Z]+/i]
+        return nil if letters.nil?
+
+        letters.upcase.each_char.reduce(0) { |acc, char| (acc * 26) + (char.ord - 64) }
+      end
+
+      def parse(xml)
+        return nil if xml.nil?
+
+        doc = Nokogiri::XML(xml)
+        doc.remove_namespaces!
+        doc
+      end
     end
   end
 end

data/lib/markdownator/{converters → renderers}/html_renderer.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 
 module Markdownator
-  module Converters
+  module Renderers
     # Walks a Nokogiri HTML node tree and renders Markdown. A focused,
     # dependency-free replacement for reverse_markdown: HTML conversion needs
     # only Nokogiri (which reverse_markdown depended on anyway).

data/lib/markdownator/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Markdownator
-  VERSION = "0.1.1"
+  VERSION = "0.1.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: markdownator
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors:
 - alexrupom
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-06-12 00:00:00.000000000 Z
+date: 2026-06-13 00:00:00.000000000 Z
 dependencies: []
 description: Markdownator converts PDF, Word, Excel, PowerPoint, EPUB, HTML, CSV,
   JSON, XML, ZIP archives and images into clean Markdown suitable for large language
@@ -33,7 +33,6 @@ files:
 - lib/markdownator/converters/docx.rb
 - lib/markdownator/converters/epub.rb
 - lib/markdownator/converters/html.rb
-- lib/markdownator/converters/html_renderer.rb
 - lib/markdownator/converters/image.rb
 - lib/markdownator/converters/json.rb
 - lib/markdownator/converters/pdf.rb
@@ -44,6 +43,7 @@ files:
 - lib/markdownator/converters/zip.rb
 - lib/markdownator/engine.rb
 - lib/markdownator/errors.rb
+- lib/markdownator/renderers/html_renderer.rb
 - lib/markdownator/result.rb
 - lib/markdownator/stream_info.rb
 - lib/markdownator/version.rb